The goal of this project is to learn the effect of neurotransmitter release by identified mammalian amacrine cells on the activity of the retinal ganglion cells. A group of methods has been developed in which the neurons of intact retinas are labeled with fluorescent nucleic acid stains. The staining makes it possible to recognize the cell bodies of subclasses of amacrine or ganglion cell under a fluorescence microscope. The neurons are then injected under visual control with lucifer yellow, which fills the dendrites. The yield of completely stained cells is an order of magnitude greater than has been previously available. Two types of experiment will be carried out. The first is anatomical in purpose; for this, dye-stained cells will be injected with lucifer yellow in lightly fixed whole mounts. An amacrine cell will be injected, and then a second, nearby ganglion or amacrine cell will be injected as well. The cells will be identified on the basis of their shapes. In some cases identification with known neurotransmitter classes will be made, by comparison of the shape of the injected amacrine cell with published immunohistochemical staining. Special attention will be paid to the amacrine cells which contain somatostatin or substance P. The spatial relations between the dendrites of the injected cell pairs will be studied by classical methods and computer aided reconstruction. The second experiment is physiological, and will study living retinas in vitro. The same methods of fluorescent staining and cell injection will be used. The activity of a retinal ganglion cell will be recorded extracellularly. A preselected amacrine cell will be penetrated by micropipette. The action of the amacrine cell upon the ganglion cell will be learned by observing the effect on ganglion cell activity of passing electrical currents into the amacrine cell. After recording, the identities and dendritic geometries of both neurons will be confirmed by lucifer injection.